Omanpx [1906686]

It's great to see more competition ideas, though I'm curious, how will you take into account the problem that plagues all energy-based competitions - higher networth, higher available energy. With attacks, you also have to consider level of the players. While everyone can get a FF 3 multiplier, base respect is determined by target level and naturally level 100s have a higher pool of high base respect targets (and likely more nw), giving them a significant advantage.

Could do level / nw brackets maybe, but it's still very tricky to do fairly imo.

Omanpx [1906686]

It's great to see more competition ideas, though I'm curious, how will you take into account the problem that plagues all energy-based competitions - higher networth, higher available energy. With attacks, you also have to consider level of the players. While everyone can get a FF 3 multiplier, base respect is determined by target level and naturally level 100s have a higher pool of high base respect targets (and likely more nw), giving them a significant advantage.

Could do level / nw brackets maybe, but it's still very tricky to do fairly imo.

Omanpx [1906686]

It's great to see more competition ideas, though I'm curious, how will you take into account the problem that plagues all energy-based competitions - higher networth, higher available energy. With attacks, you also have to consider level of the players. While everyone can get a FF 3 multiplier, base respect is determined by target level and naturally level 100s have a higher pool of high base respect targets (and likely more nw), giving them a significant advantage.

Could do level / nw brackets maybe, but it's still very tricky to do fairly imo.

Zinitheve [3001212]

Maybe simple energy banning equal to E regen. You use FHC or Can - you get 300 min ban, during which earned respect isn't included.

Zinitheve [3001212]

Maybe simple energy banning equal to E regen. You use FHC or Can - you get 300 min ban, during which earned respect isn't included.

Code [2091491]

Can't track FHC usage sadly.

But could put a daily limit on how many attacks counted. So many options and things to consider.... 

Another part of the plan is to keep it open to all, But also provide a slight bonus to those that use the discord. Can't be too much of an edge. But certainly planning a small one.

Zinitheve [3001212]

Maybe simple energy banning equal to E regen. You use FHC or Can - you get 300 min ban, during which earned respect isn't included.

Code [2091491]

Can't track FHC usage sadly.

But could put a daily limit on how many attacks counted. So many options and things to consider.... 

Another part of the plan is to keep it open to all, But also provide a slight bonus to those that use the discord. Can't be too much of an edge. But certainly planning a small one.

Zinitheve [3001212]

But you can track boosters usage, it is mostly the same 😃

Code [2091491]

A FREE brain teaser for Yall. With prize xanax if you convince me you followed it!! (Absolutely not me wanting others to check my work)

But for posterity. This is right, right? I hate time related shit...

tiksan [2383326]

It looks like you’re collating attacks from the previous day and updating aggregated scores. It should be correct based on the signatures.

 

Some tips:

Based on some function signatures, it looks like you’re using concurrent code on what appears to functions that wouldn’t be poll/epoll or async IO in general.

Use logging instead of print, dealing with journalctl on production isn’t always a great experience plus it normally is deleted upon restarting the machine.

 

Also you have some random new lines after the if statements.

tiksan [2383326]

It looks like you’re collating attacks from the previous day and updating aggregated scores. It should be correct based on the signatures.

 

Some tips:

Based on some function signatures, it looks like you’re using concurrent code on what appears to functions that wouldn’t be poll/epoll or async IO in general.

Use logging instead of print, dealing with journalctl on production isn’t always a great experience plus it normally is deleted upon restarting the machine.

 

Also you have some random new lines after the if statements.

Code [2091491]

It all gets uploaded to an AWS cluster and runs discord bots, So i make everything asynchronous... Bit of a nub and dealing with multiple bots having conflict issues has scared me into defaulting with asynchronous...

As for the logging. The AWS shit has a main log that shows all the print stuff. Super handy, saves faffing about with extra stuff. I'm sure i'll make seperate ones for errors and events eventually... But its low effort and easy for now>


And TYVM!! 

Time based stuff confuses the hell out of me... Xanax incoming!

 

tiksan [2383326]

It looks like you’re collating attacks from the previous day and updating aggregated scores. It should be correct based on the signatures.

 

Some tips:

Based on some function signatures, it looks like you’re using concurrent code on what appears to functions that wouldn’t be poll/epoll or async IO in general.

Use logging instead of print, dealing with journalctl on production isn’t always a great experience plus it normally is deleted upon restarting the machine.

 

Also you have some random new lines after the if statements.

Code [2091491]

It all gets uploaded to an AWS cluster and runs discord bots, So i make everything asynchronous... Bit of a nub and dealing with multiple bots having conflict issues has scared me into defaulting with asynchronous...

As for the logging. The AWS shit has a main log that shows all the print stuff. Super handy, saves faffing about with extra stuff. I'm sure i'll make seperate ones for errors and events eventually... But its low effort and easy for now>


And TYVM!! 

Time based stuff confuses the hell out of me... Xanax incoming!

 

Flid [2594918]

Out of interest, what sort of budget do you have with AWS usage?

 

Presently all my stuff is hosted by myself, but if I could introduce some sort of resilience then that would be lovely!

tiksan [2383326]

Not the cheapest option out there,

 

That's an understatement lol. That adds up to about $10.50 per month without including ingress/egress and other costs I assume. I pay about $13.60 per month on Hetzner for 8 vCPUs, 16 GB of RAM, 160 GB of NVME storage, and 20 TB of egress. Of course, that's bare metal so I need to manage the server myself.

tiksan [2383326]

Not the cheapest option out there,

 

That's an understatement lol. That adds up to about $10.50 per month without including ingress/egress and other costs I assume. I pay about $13.60 per month on Hetzner for 8 vCPUs, 16 GB of RAM, 160 GB of NVME storage, and 20 TB of egress. Of course, that's bare metal so I need to manage the server myself.

Code [2091491]

Still better than Google!!

I had to manage the server there and was paying more.... plus it was a headache to understand some of it....

In my defence, Been at code for under a year. All self taught... I looked into a few options after google, saw a bunch cheaper, but prioritised ease of access, reliability and GPTability.... 

And its very low effort now its set up. I just make an image of my code: Upload, click update and its all running. Then for any shit i don't want to lose on update, I throw it in my bucket. Or download shit from there or directly from the code itself. Next to zero faff.

The amount i stated was all in costs. And its been less than that for a lot of it. But very low traffic.

But yes... certainly not the cheapest, But I'm happy with my Swedish cluster and bucket (Would be a bit cheaper if i moved it from Sweden i think)

MAVRI [2402357]

Could try using something like Coolify or Easypanel (there are others that do the same, but these are the first ones that popped up into my mind) to automate most of the things during deployment. Could end up saving you a pretty penny.